Vishwamitra Evidence
Issue No #016
After three decades of commercial use, thousands of regulatory assessments and continuing post market monitoring, the evidence on approved GM foods is considerably less dramatic than the public argument around them.
By Santhosh Vishwamitra
The broad conclusion is fairly clear, but it needs to be stated carefully. Genetically modified foods that have passed national safety assessments and reached the market have not been shown to pose greater human health risks than comparable conventional foods. That does not make every GMO automatically safe, and it does not make every concern about GM agriculture a health concern. The sensible scientific position is more specific: assess the actual product, the trait, the exposure and the evidence.
That distinction matters. A question about whether a particular GM food is toxic or allergenic is not the same question as whether a farming system encourages herbicide resistance. Nor is either question settled by a political argument about labelling, seed ownership or agricultural policy.
What the major assessments actually say
The World Health Organization says that GM foods currently available on the international market have passed safety assessments and are not likely to present risks to human health. It also makes an important qualification: different GM organisms contain different genes and are produced in different ways, so individual foods should be assessed case by case. WHO also notes that no demonstrated human health effects have been attributed to consumption of the approved GM foods available in countries where they are used.
The U.S. National Academies reached a similar conclusion in its 2016 review of genetically engineered crops. After examining the available research, the committee found no substantiated evidence of a difference in human health risk between currently commercialised genetically engineered crops and conventionally bred crops. It also noted that long term epidemiological studies specifically designed around GM food consumption were limited, which is a useful reminder about what the evidence can and cannot establish.
The U.S. Food and Drug Administration takes a product based approach. Foods from genetically engineered and genome edited plants must meet the applicable food safety requirements, and the agency says the foods it has evaluated have not been more likely to cause allergic or toxic reactions than foods from traditionally bred plants. FDA has also applied the same broad food safety framework to genome edited plant varieties.
There is newer epidemiological evidence as well. A 2026 review in GM Crops & Food examined observational evidence, previous reviews and cross national disease trends. Its analysis found no consistent temporal relationship between GMO authorisation and changes in disease incidence, and concluded that the available epidemiological evidence does not consistently support associations between GMO consumption and major chronic diseases. That is useful additional evidence, but it is still a review of population level data rather than proof that no unknown health effect could ever exist.
A separate 2026 analysis looked at the global regulatory record rather than disease trends. It estimated that 4,485 risk assessments had been conducted on GM crop varieties by scientific experts in 72 countries through 2025, covering food, feed and cultivation uses. The authors reported that these assessments had not identified increased risk relative to conventional varieties. That figure should be understood as a count of regulatory risk assessments, not 4,485 independent clinical trials.
EFSA's monitoring record illustrates why the word "case by case" still matters. In its 2026 assessment of 2024 post market monitoring for maize MON 810 in Spain and Portugal, EFSA found no evidence of unexpected adverse effects on human or animal health or the environment associated with cultivation in 2024. At the same time, it identified weaknesses and uncertainties in resistance monitoring and environmental surveillance. In other words, a reassuring monitoring result did not end the monitoring programme.
What "safe" does and does not mean
In food regulation, "safe" does not mean that a product has zero conceivable risk. It means that, based on the evidence available to regulators, the product does not present an unacceptable or elevated risk under its approved conditions of use.
That distinction is easy to lose in public debate. Conventional breeding can also change the composition of a plant and, in some circumstances, alter the levels of substances relevant to nutrition, toxicity or allergenicity. That is one reason modern food safety assessment focuses on the characteristics of the resulting food rather than treating a breeding technique as a complete safety verdict.
Claim
"All GMOs are dangerous" or "All GMOs are automatically safe."
Evidence status
FALSE / MISLEADING. Major scientific and regulatory assessments do not show greater human health risk from approved commercial GM foods as a class. But safety is still assessed product by product. A blanket claim in either direction ignores how the regulatory system actually works.
Claim
Approved GM foods cause cancer, allergies or infertility.
Evidence status
NOT SUPPORTED. Major institutional assessments and the available epidemiological evidence do not establish these claims for approved commercial GM foods. The evidence does not justify turning them into established causes of disease.
Product matters more than the label "GMO"
One of the strongest ideas in the scientific and regulatory record is also one of the least exciting: the product matters.
A virus resistant papaya, insect resistant cotton, herbicide tolerant soybean and a genome edited mushroom are not the same biological product. They can have different traits, different exposure pathways and different regulatory questions. Putting all of them into one category called "GMO" tells a reader very little about the actual risk.
The National Academies made essentially this point in its 2016 review. Regulatory systems should focus on the characteristics and potential hazards of the resulting crop rather than assuming that the production method alone determines risk. That principle also explains why case by case assessment remains necessary even after decades of commercial experience.
This is not an argument that genetic engineering has no risks. It is an argument about where the risks have to be looked for. A new protein, a changed nutritional composition, an altered pesticide trait or an unexpected characteristic deserves examination. The fact that a plant was produced with a particular technology does not answer those questions by itself.
Why eating GMO food does not rewrite your genome
Claims that GMO food changes a person's genes, alters intelligence or somehow rewrites the human genome have circulated for years. They sound biological because they use the language of DNA, but the conclusion does not follow from the biology.
Humans eat DNA every day. Food DNA is exposed to digestion and cellular barriers, and the presence of DNA fragments in the gastrointestinal tract does not mean that those fragments have been incorporated into the consumer's genome. WHO has treated gene transfer as a theoretical safety question that can be considered in risk assessment, while noting the very low probability of transfer under normal dietary exposure. FDA likewise explains that DNA from GMO feed does not transfer into the DNA of animals that consume it.
The stronger claim that approved GM foods therefore rewrite human DNA has no supporting evidence. Nor is there population level evidence that eating approved GM foods reduces intelligence or produces some general hormonal effect. Those claims go well beyond what the underlying biology or health evidence shows.
Claim
GMO DNA changes your genes, alters your hormones or reduces intelligence.
Evidence status
FALSE / MISLEADING. Eating DNA does not mean that the DNA becomes part of the consumer's genome. No evidence establishes that approved commercial GM foods rewrite human DNA or cause intelligence changes through food derived genetic material.
The next regulatory question is gene editing
The regulatory debate is now moving beyond the first generation of GM crops. Gene editing allows breeders to make targeted changes to plant genomes, sometimes producing changes that could also arise through conventional breeding.
The European Union has responded with Regulation (EU) 2026/1388. The regulation entered into force on 16 July 2026, but its main provisions will apply from 17 July 2028. It creates two broad categories of plants produced using certain new genomic techniques. Category 1, commonly called NGT 1, covers plants meeting defined criteria for equivalence with conventional breeding. Category 2, commonly called NGT 2, covers plants that fall outside those criteria and remains under the stricter GMO framework. EFSA is responsible for developing scientific and administrative guidance for the new system.
The distinction is important, but it should not be oversimplified. Category 1 does not mean "every gene edited plant is treated like a conventional plant." There are eligibility conditions, verification requirements and rules for plant reproductive material. Category 2 products remain subject to risk assessment and authorisation under the existing GMO framework. NGT plants are also excluded from organic production, subject to the regulation's provisions concerning technically unavoidable presence.
This is another example of why the final characteristics of a product matter. The regulatory question is becoming less about the label attached to a technology and more about what was actually changed and what that change means.
India's different commercial reality
India provides a useful reminder that a scientific safety assessment does not automatically determine national policy.
Bt cotton remains the only GM crop approved for commercial cultivation in India. The Indian government has repeatedly described Bt cotton as the country's only commercially approved GM crop.
GM mustard has followed a different path. In October 2022, the Genetic Engineering Appraisal Committee approved environmental release of DMH 11 and its parental lines for seed production and testing, subject to conditions and existing regulatory requirements before commercial release. The matter then reached the Supreme Court. In July 2024, the Court delivered a split verdict, with the two judges taking different positions on the validity of the approval. The issue therefore cannot be described as a straightforward commercial rollout.
India has also established a regulatory route for certain genome edited plants. Its 2022 framework covers SDN 1 and SDN 2 categories that are free of exogenous introduced DNA and provides a separate regulatory pathway for their development and review.
The Indian debate therefore involves more than toxicology. It includes agricultural policy, environmental management, farmer choice, seed systems, regulatory capacity and questions about how much precaution should be built into the approval process. A finding about food safety cannot, by itself, answer all of those questions.
Food safety is not the same as environmental management
This is where many GMO arguments go off track.
A crop can be safe to eat and still create an agricultural management problem. Insect resistance can evolve. Weeds can become resistant to herbicides. Refuge requirements can be ignored. A particular farming system can create environmental pressures that are separate from the toxicity of the food itself.
EFSA's recent MON 810 assessment is a good example. The agency found no evidence of unexpected adverse effects associated with cultivation in 2024, but it also highlighted uncertainty around resistance monitoring and environmental surveillance. Those observations are not contradictory. A crop can pass a food safety assessment while still requiring active management in the field.
The same caution applies to pesticide debates. The safety of a GM trait, the safety of a particular herbicide, the amount of that herbicide used and the ecological effects of the farming system are related questions, but they are not interchangeable ones.
What remains uncertain
There are still limits to what the evidence can tell us. Long term population health research is difficult because diets and disease risks change alongside income, healthcare, smoking, pollution, occupation, lifestyle and many other factors. A population study can therefore be useful without being able to isolate every possible dietary effect.
There is also a practical distinction between products that have been consumed for years and products that have not yet entered widespread use. A future crop with a genuinely novel trait may raise questions that cannot be answered simply by pointing to the safety record of an older crop.
The same applies to genome editing. The EU's new NGT framework has not yet reached its main application date, so real world experience with commercial category 1 products under that system is still ahead. India is also still developing practical experience with its genome editing framework at scale. Those are genuine areas for continued observation.
None of this reverses the existing evidence on approved commercial GM foods. It puts that evidence in its proper place.
Editorial Judgment
The evidence supports the conclusion that approved commercial GM foods, when assessed case by case, have not been shown to carry greater human health risks than comparable conventional foods. That conclusion is strong enough to reject sweeping claims that these foods cause cancer, infertility, intelligence loss or genetic rewriting. It is not strong enough to declare every future genetically modified or gene edited product harmless before it has been assessed.
The more useful debate is therefore a narrower one. What was changed? What does the changed trait do? What will people or the environment be exposed to? What does the evidence show? And what remains uncertain? Those questions are considerably less dramatic than the usual GMO argument, but they are also much closer to how serious risk assessment actually works.
Key Takeaways
- Approved commercial GM foods that have undergone appropriate safety assessment have not been shown to present greater human health risk than comparable conventional foods.
- "Safe" does not mean zero conceivable risk. It means the available evidence does not show an unacceptable or elevated risk under the assessed conditions of use.
- Safety remains product specific. The label "GMO" does not describe one uniform biological risk.
- Food derived DNA does not provide evidence that approved GM foods rewrite the consumer's genome.
- The EU's new genomic technique framework creates category 1 and category 2 pathways, with the main provisions applying from 17 July 2028.
- India commercially cultivates Bt cotton as its only approved GM crop, while GM mustard remains tied to an unresolved regulatory and judicial process.
- Food safety, environmental effects, pesticide exposure and agricultural management should be evaluated as related but distinct questions.
- Future gene edited products still require evidence specific to the traits and products involved.
Santhosh Vishwamitra
Vishwamitra Chronicles
Sources
World Health Organization: Food, genetically modified, Q&A
National Academies of Sciences, Engineering, and Medicine (2016): Genetically Engineered Crops: Experiences and Prospects
U.S. Food and Drug Administration: Understanding New Plant Varieties and Food from New Plant Varieties
European Food Safety Authority: New Genomic Techniques materials and 2026 assessment of 2024 post market environmental monitoring for maize MON 810
Regulation (EU) 2026/1388 on plants obtained by certain new genomic techniques and their products
GM Crops & Food (2026): Genetically modified foods and human health: a comprehensive review and cross national time trend analysis
GM Crops & Food (2026): The global scientific consensus on genetically modified crop risk assessments
India GEAC, Ministry of Environment, Forest and Climate Change: GM crop approvals and DMH 11 materials
Supreme Court of India, Gene Campaign v. Union of India, 2024
Department of Biotechnology, Government of India: Guidelines and SOPs for Genome Edited Plants under SDN 1 and SDN 2 categories, 2022
Codex Alimentarius principles for the risk analysis of foods derived from modern biotechnology
Non GMO Project Standard Version 17, published 1 October 2026
Comments
Post a Comment